---
title: "Space Missions 1957/04/10 - 2022/07/29"
# author: "MG" 
date: "Modified: `r Sys.Date()`"
output: 
  flexdashboard::flex_dashboard:
    orientation: rows
    vertical_layout: fill
    social: menu 
    source_code: embed 
---

<!-- to get pdf
always_allow_html: true
output:
  pdf_document: default
-->


```{r setup, include=FALSE}
library(flexdashboard)
# to install flexdashboard
#install.packages("flexdashboard")

# flexdashboard intro
# https://pkgs.rstudio.com/flexdashboard/articles/flexdashboard.html

# for not outputting echo
knitr::opts_chunk$set(echo = FALSE, results='hide', message=FALSE, warning=FALSE)
```

```{r loadlib}
# the default timeframe of the dataframe
LAUNCH_PERIOD="10/4/1957 - 7/29/2022"
LAUNCH_PERIOD_DOT=paste0(LAUNCH_PERIOD, ".")

# most intense color 
MOST_INTENSE_COLOR="#132B43"  # dark blue
LEAST_INTENSE_COLOR="#56B1F7" # light blue
```


```{r}
# read the csv file
library(readr)
df <- read_csv('../mission-all-fields.csv')

library(dplyr)

# we will be modifying df so making a copy
df_facility <- df
# concatenate Facility, State, Area and skip NA values
df_facility$Place <- apply(df[,3:5], 1, function(x) paste(na.omit(x), collapse = ", "))
# how many NA is per column in df_facility
sapply(df_facility, function(x) sum(is.na(x)))

# ------------------
# for launches
# -----------------

# find total number of launches per place and their successes and failures
mission_status_df <- df_facility %>% group_by(Place) %>% 
  summarize(Total = n(), 
            Success = sum(MissionStatus == 1), 
            Failure = sum(MissionStatus == 2),
            PartialFailure = sum(MissionStatus == 3),
            PrelaunchFailure = sum(MissionStatus == 4),
            JustFailure = Total - Success) %>% 
  arrange(Place)

# add abbreviation for diagrams
mission_status_df$PlaceAbbrv  <- abbreviate(mission_status_df$Place, minlength = 45)

# have percentages ready for use
total_launches <- sum(mission_status_df$Total)

library(scales)

#' Compute the percentage
#' @param value Value to be presented as percentage
#' @param digi What precision, how many digits after decimal point
#' @return rounded value with digi digits after decimal point
percentage <- function(value, digi) {
  round(value*100, digits = digi)
}

# total percent success and total percent of failures
#mission_status_df$SuccessPercent <- percent(mission_status_df$Success/total_launches, accuracy = 0.1)
#mission_status_df$FailurePercent <- percent(mission_status_df$JustFailure/total_launches, accuracy = 0.1)
mission_status_df$SuccessPercent <- percentage(mission_status_df$Success/total_launches, 1)
mission_status_df$FailurePercent <- percentage(mission_status_df$JustFailure/total_launches, 1)
mission_status_df$TotalPercent <- percentage(mission_status_df$Total/total_launches, 1)
mission_status_df$RelSuccessPerc <- percentage(mission_status_df$Success/mission_status_df$Total, 1)
mission_status_df$RelFailurePerc <- percentage(mission_status_df$JustFailure/mission_status_df$Total, 1)

# find out total number of successes, failures and others
totals_df <- colSums(mission_status_df[,2:6])
```

Facilities {data-orientation=columns}
====================================================

Column {data-width=150}
-------------------------------------------------------

### Global No. of Launches

```{r, results='markup'}
valueBox(total_launches, icon="fa-rocket")
```

### No. of Launching Countries/Locations

```{r, results='markup'}
valueBox(nrow(df_facility %>% distinct(Area)), icon="glyphicon-map-marker")
```

### No. of Launching Facilities

```{r, results='markup'}
valueBox(nrow(df_facility %>% distinct(Place)), icon="fa-building")
```

Column {data-width=400}
--------------------------------------

### Histogram of no. of facilities per country. {.no-padding}

```{r, fig.width=8}
# get the number of facilities per country
countries_facilities_df <- df_facility %>% select(Area, Facility) %>% distinct(Facility, Area) %>% group_by(Area) %>% summarize(FacilityCount = n()) %>% arrange(desc(FacilityCount))

# add the percentage column
facilities_count <- sum(countries_facilities_df$FacilityCount)
countries_facilities_df$Percent <- percentage(countries_facilities_df$FacilityCount / facilities_count, 0)

hist(countries_facilities_df$FacilityCount,
     xlab = "No. of Launching Facilities", 
     ylab = "No. of Countries",
     labels=TRUE, 
     xlim = c(0, 20),
     ylim = c(0, 20),
     main = NULL)
axis(side = 1, at = seq(0,20, 1))
```

> Majority of countries had at most two launching facilities.

### Histogram of no. of launches per facility.{.no-padding}

```{r, fig.width=8}
hist(mission_status_df$Total,
     xlab = "No. of Launches", 
     ylab = "No. of Facilities",
     ylim=c(0,50),
     labels=TRUE, main = NULL)
```

> `r paste0(percentage(41/nrow(df_facility %>% distinct(Place)),0), '%')` facilities performed no more than 200 launches.

<!--
### Facilities with <= 200 launches {data-width=100}

```{r, results='markup'}
gauge(percentage(41/nrow(df_facility %>% distinct(Place)),0), min = 0, max = 100, symbol = '%'
      )
```
-->

Column {data-height=500 data-width=500}
-----------------------------

### Percent of facilities located in a country.  {.no-padding .no-title}

```{r, fig.height=5}
library(ggplot2)
ggplot(data=countries_facilities_df, aes(reorder(Area, Percent), Percent, fill=Percent)) +
  geom_bar(stat="identity") +
  geom_text(aes(label = Percent), vjust=0.5, hjust=-0.3) +
  scale_fill_continuous(high = MOST_INTENSE_COLOR, low = LEAST_INTENSE_COLOR) +
  scale_y_continuous(limits= c(0,37)) +
  xlab('Facility Location') +
  ylab('[%] of Total No. of Facilities') +
  coord_flip() +
  labs(fill='% Facilities\nLocated\nin Country', 
       title='Percent of facilities located in a country.')
```

> Rockets were launched off of vessels for Yellow Sea and Barents Sea.

By Location {data-orientation=rows data-navmenu="Launches" data-navmenu-icon="fa-rocket"}
====================================================

Row
-------------------------------------

### Locations/countries w.r.t. the global no. of launches. {.no-padding}

```{r}
# launches per location/country

launches_per_loc_df <- df_facility %>%
  group_by(Area) %>%
  summarize(Total = n(), Percent = percentage(Total/ total_launches, 1)) %>%
  arrange(desc(Total))

launches_per_loc_df %>% select (Area, Percent) %>% 
  ggplot(aes(reorder(Area, Percent), Percent, fill=Percent)) +
  geom_bar(position=position_dodge(), stat="identity") +
  geom_text(aes(label = Percent), vjust=0.5, hjust=-0.3) +
  scale_fill_continuous(high = MOST_INTENSE_COLOR, low = LEAST_INTENSE_COLOR) +
  xlab('Location/Country') +
  ylab('[%] of Total Launches') +
  ylim(0,35) +
  coord_flip() +
  labs(fill='Percent', 
       #title=paste('Locations/countries w.r.t. the global no. of launches.'),
       subtitle=paste('There were', total_launches,   'launches in total.'
                      ))
```

> Top five countries/locations accounted for over 92% of global launches.

### Top 10 launching facilities w.r.t. the number of launches. {.no-padding}

```{r}
mission_status_df %>%
  select(PlaceAbbrv, TotalPercent) %>%
  arrange(desc(TotalPercent)) %>%
  head(10) %>%
  ggplot(aes(reorder(PlaceAbbrv, TotalPercent), TotalPercent, fill=TotalPercent)) +
  geom_bar(position=position_dodge(), stat="identity") +
  geom_text(aes(label = TotalPercent), vjust=0.5, hjust=-0.3) +
  scale_fill_continuous(high = MOST_INTENSE_COLOR, low = LEAST_INTENSE_COLOR) +
  xlab('Facility') +
  ylab('[%] of Total Launches') +
  ylim(0,30) +
  coord_flip() +
  labs(fill='Percent', 
       #title='Top 10 facilities w.r.t. total no. of launches.',
       subtitle=paste('There were', total_launches,   'launches in total.'))
```

By Year {data-orientation=columns data-navmenu="Launches"}
====================================================

Column {.no-padding data-width=300}
------------------
### Global No. of Launches: `r total_launches` {.no-title data-height=115 .no-padding}

Global No. of Launches: `r total_launches`

### Space mission launch successes and failures. {.no-title data-height=400 .no-padding}

```{r, fig.width=8, fig.align='left'}
# install.packages("lessR")
library(lessR)

counts_c <- totals_df[2:5]
# Color palette
colors <- hcl.colors(length(counts_c), "Blues")
pc(counts_c, 
   radius = 1,
   hole = 0.50,
   data = counts_c,
   fill=colors,
   values_size = 1.55,
   labels_cex=1.55,
   main_cex=1.55,
   main='Space missions launch successes and failures.')
```


```{r}
# how many launches per year
#install.packages("lubridate")
library(lubridate)

# get the year from the date
# for some reason this does not work but it works in a R script
# #years_vec <- format(as.Date(df$Date, format="%Y-%m-%d"),"%Y")
years_vec <- year(mdy(df$Date))

# table() will get you the frequency table
years_df <- as_data_frame(table(years_vec))

# rename the columns names
colnames(years_df) <- c('Year', 'LaunchesCount')

# 'Year' is factor (see class(years_df$Year)), so it contains levels
# we want to convert it to numeric values and it is done via
# converting to a character and numeric
years_df$Year <- as.numeric(as.character(years_df$Year))

# add the percentage relative to the total launches
years_df$Percent <- percentage(years_df$LaunchesCount/total_launches,1)

# when is the maximum number of launches
max_number_of_launches <- years_df %>% filter( years_df$LaunchesCount == max(years_df$LaunchesCount))
# when the minimum number of launches occurred
min_number_of_launches <- years_df %>% filter( years_df$LaunchesCount == min(years_df$LaunchesCount))

# fast method for finding the second lowest; 
second_lowest <- sort(years_df$LaunchesCount,partial=2)[2]
second_min <- years_df %>% filter( years_df$LaunchesCount == second_lowest)
```

### Top 10 years with the most launches. {.no-title .no-padding data-height=400}

```{r, fig.width=8}
most_launches_per_year <- years_df %>%
  arrange(desc(LaunchesCount), desc(Year)) %>%
  head(10)
# without conversion to a character it treats is as numerical values
# and displays accordingingly at the year axis (i.e., with gaps)
most_launches_per_year$Year <- as.character(most_launches_per_year$Year)

most_launches_per_year %>% 
  ggplot(aes(reorder(Year, Percent), y=Percent, fill=Percent)) +
  geom_bar(position=position_dodge(), stat="identity") +
  geom_text(aes(label = Percent), vjust=0.5, hjust=-0.2) +
  coord_flip() +
  #scale_y_continuous(breaks = seq(from=0, to=160, by=10)) +
  #scale_x_continuous(breaks = seq(from=0, to=100, by=10))
  scale_fill_continuous(high = MOST_INTENSE_COLOR, low = LEAST_INTENSE_COLOR) +
  xlab('Year') +
  ylab('[%] of Launches Globally') +
  labs(fill='[%] of\nLaunches\nGlobally', 
       title= 'Top 10 years with the most launches a year.')
```

> In `r max_number_of_launches[1]` occurred the highest no. of launches - `r max_number_of_launches[2]`. 

Column {.no-padding data-width=600}
------------------------

### Highest and Lowest Number of Launches per Year {.no-title .no-padding}

```{r, fig.width=10, fig.height=6}
# percentages w.r.t. the global number of launches
years_df %>% 
  ggplot(aes(x=Year, y=Percent, fill=Percent)) +
  geom_bar(position=position_dodge(), stat="identity") +
  theme(axis.text.x = element_text(angle=90, vjust=.5, hjust=1, size=6)) +
  scale_fill_continuous(high = MOST_INTENSE_COLOR, low = LEAST_INTENSE_COLOR) + 
  ylab('[%] of Launches Globally') +
  labs(fill='% of\nLaunches\nGlobally', 
       title='Percentage of launches per year globally.')

# for absolute numbers you can use this
# years_df %>% 
#  ggplot(aes(x=Year, y=LaunchesCount, fill=LaunchesCount)) +
#  geom_bar(position=position_dodge(), stat="identity") +
#  theme(axis.text.x = element_text(angle=90, vjust=.5, hjust=1, size=6)) +
#  scale_fill_continuous(high = MOST_INTENSE_COLOR, low = LEAST_INTENSE_COLOR) + 
#  ylab('No. of Launches') +
#  labs(fill='Number of\nLaunches', 
#       title=paste(
#         'The number of launches per year globally for the period', LAUNCH_PERIOD_DOT))
```

> No. of launches was the highest (~ 100 per year) in 1966-1978, and 2016-2022. Lowest no. of launches per year was performed in 2001-2015 and it ranged from 
`r min(years_df[years_df$Year >= 2001 & years_df$Year <= 2015,]$LaunchesCount)`
(`r min(years_df[years_df$Year >= 2001 & years_df$Year <= 2015,]$Percent)`% of global launches) to `r max(years_df[years_df$Year >= 2001 & years_df$Year <= 2015,]$LaunchesCount)` (`r max(years_df[years_df$Year >= 2001 & years_df$Year <= 2015,]$Percent)`% of global launches). 

Rockets {data-orientation=columns}
====================================================

Column {data-width=280}
---------------------------


```{r}
# how many and what type of rockets are used
# how many times each rocket was used and what is its status (retired or not)
# how long it served

# first deal with the date and convert it to the right format
# for some reason the code that works fine in the R script
# does not work in a R Markdown chunk so here baby steps
rockets_df <- df %>%
  select(Rocket, RocketStatus, Date) 
# convert the Date column from character to the Date
rockets_df$Date <- mdy(rockets_df$Date)

# now deal with other stuff related to a date column
# the OperationInYears - is an approximate number (the year is assumed to
# have 365 days)
rockets_df <- rockets_df %>%
  group_by(Rocket, RocketStatus) %>% 
  summarize(LaunchCount = n(), FirstLaunch = min(Date), LastLaunch = max(Date),
            OperationInDays = length(seq(from = min(Date), to = max(Date), by = 'day')) - 1,
            OperationInYears = round(OperationInDays / 365, digits = 1)) %>% 
  arrange(desc(LaunchCount), desc(Rocket))

rockets_df$CountPerc <- percentage(rockets_df$LaunchCount / sum(rockets_df$LaunchCount), 1)

# hack to make nice labels in PieChart - I need to add a label
# otherwise I have 0 and 1, so here we go
# add the column with labels
rockets_df$RocketStatusName <- with(rockets_df, ifelse(RocketStatus == 0, "Retired", "Active"))
```


### Rockets: Active vs. Retired {.no-padding .no-title data-height=400} 

```{r}
# how many rockets are active vs retired
PieChart(RocketStatusName, data=rockets_df,
  fill=c("red","lightblue"),
   values_size = 1.5,
   values = "%",
   labels_cex = 1.5,
   cex = 1.5,
  hole = 0.4,
   main="Active vs. retired rockets as on 7/29/2022."
)
```

### Rockets {.no-padding .no-title data-height=500}
  
```{r, fig.height=5}
hist(rockets_df$LaunchCount, xlab = "No. of Launches", 
     ylab = "Rocket Count",
     xlim=c(0,500), ylim=c(0,400), labels=TRUE, main =  'Histogram of no. of launches per rocket.'
)
```

> `r percentage(359/nrow(rockets_df), 0)`% rockets launched $\le$ 50 times.


Column {data-width=400}
----------------------------------------

### Rockets Count {.no-title .no-padding data-height=250}

No. of rockets: `r nrow(rockets_df)`

No. of launches: `r total_launches`


### Top 10 Most Popular Rockets {.no-padding .no-title data-height=800}

```{r}
# top 10 most popular rockets
top10_rockets_df <- rockets_df %>% 
  select(Rocket, LaunchCount, CountPerc, RocketStatusName, FirstLaunch, LastLaunch, OperationInYears) %>% 
  arrange(desc(LaunchCount)) %>% 
  head(10) 
```

```{r, fig.height=6}
top10_rockets_df %>% 
  ggplot(aes(reorder(Rocket, CountPerc), y=CountPerc)) +
  geom_bar(position=position_dodge(), stat="identity", aes(fill=factor(RocketStatusName), group=1), show.legend = TRUE) +
  theme(axis.text.x = element_text(angle=0, vjust=.5, hjust=1)) +
  scale_fill_discrete(labels = c("Active", "Retired")) +
  #scale_fill_continuous(high = MOST_INTENSE_COLOR, low = LEAST_INTENSE_COLOR) +
  ylab('[%] of Total No. of Launches') +
  xlab('Rocket') + 
  labs(fill="Rocket Status", 
       title='Top 10 most popular rockets of all times.') +
  coord_flip()
```

> Rocket `r top10_rockets_df$Rocket[1]` was the most popular one with 
`r top10_rockets_df$LaunchCount[1]` launches (`r top10_rockets_df$CountPerc[1]`% of total launches).

Column {data-width=500}
-------------------------

### Top 10 Longest Serving Rockets {.no-title .no-padding data-height=400}
```{r}
# top 10 rockets serving the longest
top10_longest_rockets_df <- rockets_df %>% 
  select(Rocket, LaunchCount, CountPerc, RocketStatusName, FirstLaunch, LastLaunch, OperationInYears) %>% 
  arrange(desc(OperationInYears)) %>% 
  head(10) 
```


```{r, fig.width=10}
hist(rockets_df$OperationInYears, xlab = "Operation in Years", 
     ylab= "Rocket Count",
     xlim=c(0,50), ylim=c(0,250), labels=TRUE, breaks=9,
     main = 'Histogram of operational rocket lifespan.')
```

> `r percentage(237/nrow(rockets_df), 0)`% rockets operated for $\le$ 5 years.


### Longest Serving Rockets {.no-title .no-padding data-height=600}


```{r, fig.width=9}
top10_longest_rockets_df %>% 
  ggplot(aes(reorder(Rocket, OperationInYears), y=OperationInYears)) +
  geom_bar(position=position_dodge(), stat="identity", aes(fill=factor(RocketStatusName), group=1), show.legend = TRUE) +
  theme(axis.text.x = element_text(angle=0, vjust=.5, hjust=1)) +
  scale_fill_discrete(labels = c("Active", "Retired")) +
  #scale_fill_continuous(high = MOST_INTENSE_COLOR, low = LEAST_INTENSE_COLOR) +
  ylab('Years') +
  xlab('Rocket') +
  labs(fill="Rocket Status", 
       title='Rockets with the longest operational lifespan.') +
  coord_flip()
```

> `r top10_longest_rockets_df[1,][1]` was operational for `r top10_longest_rockets_df[1,][7]` years (from `r top10_longest_rockets_df[1,][5]` to `r top10_longest_rockets_df[1,][6]`). Longest serving rocket still active has been `r top10_longest_rockets_df[3,][1]` (`r top10_longest_rockets_df[3,][7]` years).


About {data-orientation=rows}
====================================================
       
       
       
       
       
       

Row 
---------------------------
       
### About {.no-title}

#### Data
  The data for this study is located at Maven Analytics [https://app.mavenanalytics.io/datasets] "Space Missions" and
  contains 4,630 records regarding space missions around the world for the period 10/4/1957 - 07/29/2022.

#### License
I reached out to the staff at Maven Analytics and they said that the dataset can be freely used.

#### Data Credibility
I performed some data credibility check by checking and confirming from other sources some randomly selected observations.
       
#### Data Processing
- I performed cleaning of the data. 
- Some data is missing, especially the price field.
- Data has been slightly reformatted compared to the original data for the purpose of the analysis.

#### Thanks 
       
Thanks for looking at my work.
       
Comments on how it can be done better are always welcome.